Eurobites: World is sleepwalking toward a Big Tech AI cartel, warns UK regulator

Artificial intelligence may have everyone and their uncle frothing at the mouth in excitement, but the UK's Competition and Markets Authority is sounding the klaxon of caution. It believes the usual Big Tech suspects are beginning to exert an unhealthy strangehold over the so-called foundation models that underpin AI and uses an update paper to air some thoughts about how to lessen the risks of creating what is effectively an AI cartel. "When we started this work, we were curious. Now, with a deeper understanding and having watched developments very closely, we have real concerns," says Sarah Cardell, CEO of the CMA. "By stepping up our merger review, we hope to gain more clarity and that clarity will also benefit the businesses themselves."

Another UK regulatory body, the Information Commissioner's Office, also has its concerns about how this whole AI business is developing, but it's primarily worried about the data privacy angle. John Edwards, UK Information Commissioner, said: "In a world where misinformation is growing, we cannot allow misuse of generative AI to erode trust in the truth. Organizations developing and deploying generative AI must comply with data protection law – including our expectations on accuracy of personal information." In other words, be careful what you scrape from the bottom of the online barrel.

No such concerns, apparently, from UK Prime Minister Rishi Sunak, who can't wait till he meets his South Korean opposite number, President Yoon Suk Yeol, at the AI Seoul Summit on May 21-22, which is a follow-up to the UK-hosted AI Safety Summit in December. (Though note that the "safety" bit has now been quietly dropped from the event's name. Bit of a downer.) Day one will see a "virtual leaders' session" co-chaired by Sunak and Yoon; day two will see an in-person meeting of digital ministers.

Away from the talking shops and the gladhanding, Orange is injecting AI into its network operations, via Augtera's Network AI platform. By adding Augtera's smarts into its NOC tools Orange hopes to reduce the number of network alarms displayed by 70% and better anticipate potential network incidents before too much havoc is wreaked. The integration of Augtera software begins this month and Orange says it will be rolled out globally across all its networks by the end of the year.

Swisscom and Ericsson are extending their existing contract by another three years, the operator hoping that the addition of  AI, automation and related Ericsson fairy-dust will, in the words of Swisscom CTIO Gerd Niehage, "turn Switzerland's best network into its smartest one." As part of the plan, it is envisaged that Ericsson and Swisscom employees will collaborate in mixed teams.